OpenAI API: Fine-Tuning
48mGeneral2024-04-23
Authors
Morten Rand-Hendriksen
Senior Staff Instructor, Speaker, Web Designer, and Software Developer
Course details
Fine-tuning GPT models allows you to submit examples of the types of responses the AI system should provide based on user input. This is a great way of ensuring the system speaks in a voice and tone of your choosing, and eliminates the need for providing a description of the agent before each prompt. OpenAI’s API enables you to submit your own fine-tuning training data. In this course, Senior Staff Instructor and AI whisperer Morten Rand-Hendriksen shows you how to fine-tune OpenAI’s GPT models by uploading your own data to create a unique custom model. This course takes you through the process of preparing your data, submitting the data for fine-tuning through the OpenAI API, and using the fine-tuned model for regular interactions with the AI API. Fine-tuning GPT provides better performance, enables you to shorten your prompts, and keeps the AI system on track and on target with your business content and context.
Learning objectives
Prepare your data for fine-tuning
Set up a fine-tuned model
Use a fine-tuned model through the API
Update an existing fine-tuned model
Compare performance between models
Learning objectives
Prepare your data for fine-tuning
Set up a fine-tuned model
Use a fine-tuned model through the API
Update an existing fine-tuned model
Compare performance between models
Skills covered
OpenAI APIAPIsOpenAIGenerative AIArtificial Intelligence (AI)Software DevelopmentOne-Off
Concepts
0. Introduction
- 01 - Fine-tuning custom models with the OpenAI API
- 02 - Exercise files and where to find them
1. Fine-Tuning Basics
- 03 - When to create a fine-tuned model
- 04 - Creating and formatting training data
- 05 - Testing the training data
- 06 - Creating a fine-tuning job in the playground
- 07 - Using a fine-tuned model in the playground
- 08 - Testing epoch-based checkpoints
2. Advanced Tools and Features
- 09 - Fine-tuning through the API
- 10 - Uploading training data to the API
- 11 - Creating a fine-tuning job through the API
- 12 - Retrieving a fine-tuning job and checking the status
- 13 - Getting the model name once the job is completed
- 14 - Using the fine-tuned model through the API
- 15 - Cancelling a fine-tuning job
3. Further Info
- 16 - Additional notes on fine-tuning